TMEM165 deficiencies lead to one of the congenital disorders of glycosylation (CDG), a group of inherited diseases where the glycosylation process is altered. We recently demonstrated that the Golgi glycosylation defect due to TMEM165 deficiency resulted from a Golgi manganese homeostasis defect and that Mn2+ supplementation was sufficient to rescue normal glycosylation. In the present paper, we highlight TMEM165 as a novel Golgi protein sensitive to manganese. When cells were exposed to high Mn2+ concentrations, TMEM165 was degraded in lysosomes. Remarkably, while the variant R126H was sensitive upon manganese exposure, the variant E108G, recently identified in a novel TMEM165-CDG patient, was found to be insensitive. We also showed that the E108G mutation did not abolish the function of TMEM165 in Golgi glycosylation. Altogether, the present study identified the Golgi protein TMEM165 as a novel Mn2+-sensitive protein in mammalian cells and pointed to the crucial importance of the glutamic acid (E108) in the cytosolic ELGDK motif in Mn2+-induced degradation of TMEM165.

Introduction

Manganese is a trace element essential for life. It is involved in the catalytic domain of many enzymes such as Golgi glycosyltransferases, mitochondrial enzymes, and DNA and RNA polymerases. Regulation of its homeostasis is therefore particularly important. Manganese overexposure has been shown to induce neurological symptoms that can result in a Parkinson-like disorder called manganism [13]. On the contrary, a decrease in cellular Mn2+ has recently been shown to cause congenital disorders of glycosylation (CDG). Mutations in SLC39A8, a putative plasma membrane manganese transporter, lead to severe glycosylation defects [4]. We recently reported that TMEM165 deficiency was also linked with Golgi Mn2+ homeostasis defect [5].

Although progress has been made in identifying cellular Mn2+ transporters in mammals, the mechanisms of Mn2+ homeostasis are still unclear. Several different transporters have been involved in manganese transport mechanisms, including the divalent metal transporter 1 (DMT1/NRAMP2/SLC11A2) [1,6], NRAMP1 [7], the transferrin receptor, and the transporters SLC30A10/ZNT8 [8], SLC39A8/ZIP8 [4], and SLC30A14/ZIP14 [9,10]. At the cellular level, most of these transporters are localized at the plasma membrane and/or in endosomes. The secretory pathway consisting of the ER, the Golgi and associated vesicles is also crucial in regulating cellular Mn2+ homeostasis. In addition, the secretory pathway requires luminal Mn2+ concentration for quality control, proper targeting and processing of proteins. Current knowledge supports that this supply is realized via the action of SPCA1 (secretory pathway Ca-ATPase 1: ATP2C1) and SPCA2 (ATP2C2). SPCA1 is ubiquitously expressed and mediates the import of Ca2+/Mn2+ into the Golgi lumen [11,12]. The tissue expression of SPCA2 is more restricted. However, the importance of the dual transport function in cellular processes is not yet completely deciphered [11]. Overexpression of SPCA1 has been shown to facilitate Mn2+ accumulation into the Golgi [13] and it was thus proposed that SPCA1 was a way to detoxify cytosolic Mn2+ accumulation by sequestering it into the secretory pathway.

In 2012, we identified TMEM165 as a novel Golgi transmembrane protein causing CDG [14]. It belongs to an uncharacterized family of membrane proteins named UPF0016 (Uncharacterized Protein Family 0016; Pfam PF01169). We recently demonstrated that the observed Golgi glycosylation defect resulted from Golgi Mn2+ homeostasis impairment [5]. Based on these results, we hypothesized that TMEM165 could be a novel Golgi Mn2+ transporter. As studies of Mn2+ homeostasis in yeasts have indicated that most of the proteins involved in regulating intracellular Mn2+ concentrations are differentially targeted and/or degraded in response to Mn2+, TMEM165 was tested.

The aim of the present study was to decipher the impact of high extracellular Mn2+ concentrations on the subcellular localization and stability of TMEM165. The present study demonstrates that high concentrations of extracellular Mn2+ lead to a rapid lysosomal degradation of TMEM165. We identified the glutamic acid (E108) in the highly conserved motif ELGDK, oriented toward the cytosol, as being crucial in the Mn-induced degradation of TMEM165.

Results

TMEM165 is rapidly and specifically degraded in response to Mn2+

Our previous work highlighted a link between TMEM165 and Golgi Mn2+ homeostasis [5]. As many proteins involved in regulating intracellular Mn2+ homeostasis are directly impacted in their stability by cellular Mn2+ homeostasis changes, the effect of Mn2+ on TMEM165 was tested. For this, a concentration of 500 µM of MnCl2 was first used for different times and the stability of TMEM165 was assessed both by western blot and immunofluorescence experiments. We observed that in response to Mn2+, TMEM165 levels were significantly reduced (Figure 1A,B). Interestingly, the same sensitivity to Mn2+ was observed for Gdt1p, the yeast ortholog of TMEM165 (Figure 1D). The effects of other ions were also tested (Supplementary Figure S1). Remarkably, we observed that TMEM165 degradation only occurred after MnCl2 exposure, pointing to the specificity of TMEM165 for Mn2+. As GPP130 has also been shown to be sensitive to high Mn2+ concentrations, we compared its time course degradation with TMEM165 (Figure 1A–C). Quantification indicated that TMEM165 loss exceeded 95% after 8 h of Mn2+ treatment, while only a 40% decrease was seen for GPP130. To further tackle the minimal Mn2+ concentration able to induce a loss of TMEM165, we analyzed the stability of TMEM165 with low MnCl2 concentrations (1–50 µM). While 100 µM MnCl2 was sufficient to induce GPP130 degradation [15], our results showed that 1–25 µM of Mn2+ was already sufficient to induce a destabilization of TMEM165 (Figure 1B). Altogether these results indicate that TMEM165, compared with GPP130, is more sensitive to manganese and probably suggests the existence of different degradation mechanisms in response to Mn2+. The impact of Mn2+ on TMEM165 was also seen by immunofluorescence where a decrease in TMEM165 fluorescence associated with Golgi was seen (Supplementary Figure S2A,B).

TMEM is rapidly degraded in response to Mn2+.

Figure 1.
TMEM is rapidly degraded in response to Mn2+.

(A) Steady-state cellular level of TMEM165 and GPP130. HeLa cells were treated with MnCl2 500 µM for 0–8 h. Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 and GPP130 protein levels after normalization with actin [Number of experiments (N) = 2; ***P-value < 0.001]. (B) Steady-state cellular level of TMEM165. HEK293 cells were treated with MnCl2 from 0 to 50 µM for 36 h. Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Lower panel shows the quantification of TMEM165 protein levels after normalization with actin (N = 2; ***P-value < 0.001). (C) Steady-state cellular level of GPP130 in the same experimental conditions as described in (B). (D) Gdt1Δ yeasts expressing Gdt1p-Myc were cultured in the absence or presence of 1 mM MnCl2. Yeast lysates were then subjected to SDS–PAGE and western blot analysis with the indicated antibodies. Right panel shows the quantification of the Gdt1p-Myc protein levels (N = 2; ***P-value < 0.001).

Figure 1.
TMEM is rapidly degraded in response to Mn2+.

(A) Steady-state cellular level of TMEM165 and GPP130. HeLa cells were treated with MnCl2 500 µM for 0–8 h. Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 and GPP130 protein levels after normalization with actin [Number of experiments (N) = 2; ***P-value < 0.001]. (B) Steady-state cellular level of TMEM165. HEK293 cells were treated with MnCl2 from 0 to 50 µM for 36 h. Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Lower panel shows the quantification of TMEM165 protein levels after normalization with actin (N = 2; ***P-value < 0.001). (C) Steady-state cellular level of GPP130 in the same experimental conditions as described in (B). (D) Gdt1Δ yeasts expressing Gdt1p-Myc were cultured in the absence or presence of 1 mM MnCl2. Yeast lysates were then subjected to SDS–PAGE and western blot analysis with the indicated antibodies. Right panel shows the quantification of the Gdt1p-Myc protein levels (N = 2; ***P-value < 0.001).

We previously demonstrated that TMEM165 could be found at the plasma membrane [14]. To assess the impact of Mn2+ on the plasma membrane targeted form of TMEM165, surface protein biotinylation was performed in the absence and presence of 500 µM of MnCl2. Interestingly biotin-labeled cell surface TMEM165 displayed the same sensitivity to Mn2+ as the cellular TMEM165. This either suggests that the Mn2+-induced degradation mechanism is not only dedicated to the Golgi pool of TMEM165 or that less TMEM165 traffics to the plasma membrane from the Golgi when the Golgi pool of TMEM165 has been depleted upon excess manganese exposure (Figure 2A). Previous studies have also demonstrated that in yeast, high environmental Ca2+ concentrations in gdt1Δ led to strong N-glycosylation deficiencies [5]. The impact of Ca2+ on Mn2+-induced degradation of TMEM165 was then assessed by western blot and immunofluorescence (Figure 2B,C). Ca2+ alone had no significant effect on the stability of TMEM165. However, its combined presence with Mn2+ clearly decreased the Mn2+-induced degradation of TMEM165 (80% decrease for the Mn2+ treatment alone compared with 40% decrease for both Ca2+ and Mn2+; Figure 2B). This was confirmed by confocal microscopy (Figure 2C).

Plasma membrane TMEM165 is also degraded by Mn2+ and Ca2+ compete with Mn2+ for TMEM165 degradation.

Figure 2.
Plasma membrane TMEM165 is also degraded by Mn2+ and Ca2+ compete with Mn2+ for TMEM165 degradation.

(A) Cell surface biotinylation was performed in the absence and presence of MnCl2 500 µM. Samples were prepared as described in the materials and methods section and subjected to SDS–PAGE and western blot with the indicated antibodies. Lower panel shows the quantification of TMEM165 protein levels. The plasma membrane panel is separated from the cellular panel as we had to expose films longer to reveal the bands. (B) HEK293 cells were incubated with 500 µM MnCl2 and/or 2 mM CaCl2 for 4 h, and then subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 protein levels after normalization with actin (N = 2; ***P-value < 0.001). (C) HEK293 cells were incubated with 500 µM MnCl2 and/or 2 mM CaCl2 for 4 h, and then fixed and labeled with antibodies against TMEM165 and GM130 before confocal microscopy visualization (N = 2; ***P-value < 0.001). Right panel shows the quantification of the associated TMEM165 fluorescence intensity (N = 2; n = 50; ***P-value < 0.001).

Figure 2.
Plasma membrane TMEM165 is also degraded by Mn2+ and Ca2+ compete with Mn2+ for TMEM165 degradation.

(A) Cell surface biotinylation was performed in the absence and presence of MnCl2 500 µM. Samples were prepared as described in the materials and methods section and subjected to SDS–PAGE and western blot with the indicated antibodies. Lower panel shows the quantification of TMEM165 protein levels. The plasma membrane panel is separated from the cellular panel as we had to expose films longer to reveal the bands. (B) HEK293 cells were incubated with 500 µM MnCl2 and/or 2 mM CaCl2 for 4 h, and then subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 protein levels after normalization with actin (N = 2; ***P-value < 0.001). (C) HEK293 cells were incubated with 500 µM MnCl2 and/or 2 mM CaCl2 for 4 h, and then fixed and labeled with antibodies against TMEM165 and GM130 before confocal microscopy visualization (N = 2; ***P-value < 0.001). Right panel shows the quantification of the associated TMEM165 fluorescence intensity (N = 2; n = 50; ***P-value < 0.001).

Lysosomal degradation of TMEM165

As shown by Mukhopadhyay et al. [16], 500 µM MnCl2 treatment induces rapid redistribution of GPP130 in vesicles before their lysosomal degradation. At the opposite of GPP130, no redistribution from the Golgi to peripheral punctate structures was observed for TMEM165 in response to high Mn2+ concentration (Supplementary Figure S2C). This absence of vesicles could be explained by an extremely fast degradation. To test this hypothesis, the stability of TMEM165 in response to Mn2+ was studied by immunofluorescence in the presence of chloroquine, a lysosomal inhibitor (Figure 3A). Although cells treated with Mn2+ alone showed a dramatic loss of TMEM165, those treated in the presence of chloroquine exhibited an accumulation of TMEM165 in punctate structures (Figure 3A). Immunofluorescence experiments with LAMP2, a lysosomal marker, confirmed the presence of TMEM165 in LAMP2-positive structures in chloroquine- and Mn2+-treated cells. The level of co-localization was determined using Manders' overlap coefficient and revealed a co-localization of TMEM165 with LAMP2 (72 ± 9%). Same experiments were also done with EEA1 as a specific marker of early endosomes and the quantification revealed no significant co-localization of TMEM165 with EEA1 (7 ± 1%; data not shown). This result shows that TMEM165 is specifically targeted to lysosomal degradation followed Mn2+ exposure. To confirm the lysosomal Mn2+-induced degradation of TMEM165, an immunoblotting experiment was also performed. As shown in Figure 3B, the Mn-induced degradation of TMEM165 was completely blocked by chloroquine. As chloroquine is known to both shut down endosomal trafficking and inhibit lysosomal proteases, we also tested the effects of leupeptin, a lysosomal protease inhibitor, on the stability of TMEM165 in response to Mn2+ (Supplementary Figure S3). The experiment confirmed the localization of TMEM165 in LAMP2-positive structures as a co-localization of TMEM165 with LAMP2 (62 ± 1%) and the absence of co-localization of TMEM165 with EEA1 (3 ± 1%; data not shown) was observed.

TMEM165 is targeted to lysosomal degradation after Mn2+ exposure.

Figure 3.
TMEM165 is targeted to lysosomal degradation after Mn2+ exposure.

(A) HeLa cells were treated for 8 h with 500 µM MnCl2 and/or 100 µM chloroquine, and fixed and labeled with antibodies against TMEM165 (upper panels) and LAMP2 (middle panels) before confocal microscopy visualization. (B) Western blot analysis of the same experiment described in (A). Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of the TMEM165 protein level after normalization with actin (N = 2; ***P-value < 0.001).

Figure 3.
TMEM165 is targeted to lysosomal degradation after Mn2+ exposure.

(A) HeLa cells were treated for 8 h with 500 µM MnCl2 and/or 100 µM chloroquine, and fixed and labeled with antibodies against TMEM165 (upper panels) and LAMP2 (middle panels) before confocal microscopy visualization. (B) Western blot analysis of the same experiment described in (A). Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of the TMEM165 protein level after normalization with actin (N = 2; ***P-value < 0.001).

The amino acid E108 of the ELGDK motif is involved in Mn2+-induced degradation of TMEM165

To gain more insight into the TMEM165 Mn2+-induced degradation mechanism, we wondered whether the reported missense mutations (pE108G and pR126H) found in TMEM165-deficient CDG patients could impact the TMEM165 Mn2+ sensitivity. To test the putative role of these mutations in TMEM165 Mn2+ responsiveness, immunofluorescence and western blot experiments were performed in the absence and presence of MnCl2 (Figure 4A,B). The missense mutation c.323 A>G (p.E108G) found in two newly TMEM165 deficient siblings was first tested [17]. As observed for wt-TMEM165, the mutant form is Golgi-localized in fibroblasts for the two siblings. This indicates that the mutation does not disturb the subcellular localization of the mutated form of TMEM165 (Figure 4B). The impact of Mn2+ treatment was then investigated during an 8 h time course by western blot and immunofluorescence experiments (Figure 4A,B). As expected, the wild-type (wt) TMEM165 was very sensitive to Mn2+ exposure (Figure 4A). However, the mutated form of TMEM165 (E108G) remained stable (Figure 4A,B). No changes were observed in localization or stability by immunofluorescence. Quantification of the western blot results indicated that wt-TMEM165 loss exceeded 95% at the 6 h time point, while only 20% loss was observed for the mutated form p.E108G. To demonstrate the distinctive feature of this mutation, fibroblasts from another TMEM165-CDG patient, carrying the R126H mutation, were also tested for Mn2+ sensitivity by western blot and confocal microscopy (Figure 5). Although the steady-state level of TMEM165, compared with control fibroblasts, is lower in the R126H patients' fibroblasts, our results highlighted that this mutation did not prevent the Mn2+-induced TMEM165 (R126H) degradation. It is also important to note that its localization is not altered, neither at the steady-state level nor after chloroquine and Mn2+ exposure (Supplementary Figure S5). After Mn2+ and chloroquine exposure, the co-localization between TMEM165 and LAMP2/EEA1 was determined for both control and patient fibroblasts. For control fibroblasts, we observed a co-localization of TMEM165 with LAMP2 (59 ± 5%) and no significant co-localization of TMEM165 with EEA1 (6 ± 1%). For patient fibroblasts (R126H), the results were the same. A co-localization of TMEM165 with LAMP2 (58 ± 11%) but no significant co-localization of TMEM165 with EEA1 (5 ± 1%) was observed. Interestingly, the western blot results showed that the R126H variant is stabilized upon Mn2+ and chloroquine exposure (Supplementary Figure S5C). This clearly demonstrates that this allele is constitutively able to traffic to the lysosomes upon Mn2+ exposure. In summary, our results support the evidence that the glutamic acid (E) of the highly conserved ELGDK motif is crucial in mediating the lysosomal degradation of TMEM165 in response to Mn2+.

The glutamic acid (E1108) in the ELDGK motif is crucial for Mn2+ sensitivity.

Figure 4.
The glutamic acid (E1108) in the ELDGK motif is crucial for Mn2+ sensitivity.

(A) Healthy skin fibroblasts (upper left) and patients’ skin fibroblasts (lower left) carrying E108G mutation were treated with 500 µM MnCl2 for 0–8 h. Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 protein levels after normalization with actin (N = 2; ***P-value < 0.001). (B) Healthy skin fibroblasts and patients’ skin fibroblasts carrying E108G mutation were treated with 500 µM MnCl2 for 0, 4, and 8 h. Cells were then fixed and labeled with antibodies against TMEM165 and GM130 before confocal microscopy visualization (N = 2; ***P-value < 0.001). Lower panel shows the quantification of the associated TMEM165 fluorescence intensity (N = 2; n = 30; ***P-value < 0.001). (C) HEK293 control cells and HEK293 KO TMEM165 cells were transfected with empty-vector, wt or E108G plasmid for 36 h. Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies.

Figure 4.
The glutamic acid (E1108) in the ELDGK motif is crucial for Mn2+ sensitivity.

(A) Healthy skin fibroblasts (upper left) and patients’ skin fibroblasts (lower left) carrying E108G mutation were treated with 500 µM MnCl2 for 0–8 h. Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 protein levels after normalization with actin (N = 2; ***P-value < 0.001). (B) Healthy skin fibroblasts and patients’ skin fibroblasts carrying E108G mutation were treated with 500 µM MnCl2 for 0, 4, and 8 h. Cells were then fixed and labeled with antibodies against TMEM165 and GM130 before confocal microscopy visualization (N = 2; ***P-value < 0.001). Lower panel shows the quantification of the associated TMEM165 fluorescence intensity (N = 2; n = 30; ***P-value < 0.001). (C) HEK293 control cells and HEK293 KO TMEM165 cells were transfected with empty-vector, wt or E108G plasmid for 36 h. Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies.

TMEM165 Mn2+-induced degradation also occurs in fibroblasts carrying R126H mutation.

Figure 5.
TMEM165 Mn2+-induced degradation also occurs in fibroblasts carrying R126H mutation.

(A) Control skin fibroblasts and patient skin fibroblasts carrying R126H mutation were treated with 500 µM MnCl2 for 0–8 h and then fixed and labeled with antibodies against TMEM165 and GM130 before confocal microscopy visualization. Right panel shows the quantification of the associated TMEM165 fluorescence intensity [N = 2; number of cells (n) = 30; ***P-value < 0.001]. (B) Steady-state cellular level of TMEM165. Fibroblasts were treated as described in (A). Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 protein levels after normalization with actin (N = 2; ***P-value < 0.001).

Figure 5.
TMEM165 Mn2+-induced degradation also occurs in fibroblasts carrying R126H mutation.

(A) Control skin fibroblasts and patient skin fibroblasts carrying R126H mutation were treated with 500 µM MnCl2 for 0–8 h and then fixed and labeled with antibodies against TMEM165 and GM130 before confocal microscopy visualization. Right panel shows the quantification of the associated TMEM165 fluorescence intensity [N = 2; number of cells (n) = 30; ***P-value < 0.001]. (B) Steady-state cellular level of TMEM165. Fibroblasts were treated as described in (A). Total cell lysates were prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 protein levels after normalization with actin (N = 2; ***P-value < 0.001).

To determine whether the E108G mutation could also affect the function of TMEM165, the glycosylation status of LAMP2 was assessed in TMEM165 KO HEK293 cells generated by CRISPR-Cas9 (Figure 4C and Supplementary Figure S4). Both the expression of the wt-TMEM165 and the E108G mutant complemented the observed glycosylation defect. Compared with the expression of the wt TMEM165, the expression of the E108G mutant in rescuing the LAMP2 glycosylation is less efficient. This result suggests, nevertheless, that the E108G mutant remains functional and that the activity of TMEM165 then appears independent of the Mn2+-induced degradation mechanism.

Validation of a predicted topology of TMEM165

Human TMEM165 encodes a 7-transmembrane spanning protein of 324 amino acids. To validate a predicted topology of TMEM165 and thus the orientation of the ELGDK motif, we used the two available commercial antibodies against TMEM165, each recognizing two differentially oriented epitopes: the Sigma antibodies recognizing the immunogen sequence (aa176–aa229) and the antibodies provided by Thermo Fischer directed against the immunogen sequence (aa17–aa45; Figure 6A). The topology was determined by selective membrane permeabilization and immunofluorescence analysis (Figure 6B). Under conditions that allowed antibody access to all cellular compartments, both epitopes were detectable and showed co-localization with the Golgi marker GM130 (Figure 6B). Selective permeabilization of the plasma membrane with low concentrations of digitonin allowed visualization of the cytosolic epitope only recognized by the Sigma antibody. On the basis of these results, we can propose a model where the loop encompassing the aa 176–229 is cytosolic and where the ELGDK motif is facing the cytosol (Figure 6A).

TMEM165 topology.

Figure 6.
TMEM165 topology.

(A) Representation of TMEM165 predicted topology. The two anti-TMEM165 antibodies (Sigma–Aldrich and Thermo Fisher Scientific) depicted here recognize two different parts of the protein. The Sigma ones recognize the cytoplasmic loop between the fourth and fifth transmembrane domains. The Thermo ones recognize the short luminal loop between the first and the second transmembrane domain. The red double arrows show the predicted location of the signal sequence cleavage. The first TMD is depicted in white and with a dotted line border as it can be absent from the mature protein. (B) Cells were fixed with PAF 4% and treated as described in the materials and methods section. Selective permeabilization was done by using Triton X-100 or digitonin. Cells were labeled with antibodies against TMEM165 and GM130 before confocal microscopy visualization.

Figure 6.
TMEM165 topology.

(A) Representation of TMEM165 predicted topology. The two anti-TMEM165 antibodies (Sigma–Aldrich and Thermo Fisher Scientific) depicted here recognize two different parts of the protein. The Sigma ones recognize the cytoplasmic loop between the fourth and fifth transmembrane domains. The Thermo ones recognize the short luminal loop between the first and the second transmembrane domain. The red double arrows show the predicted location of the signal sequence cleavage. The first TMD is depicted in white and with a dotted line border as it can be absent from the mature protein. (B) Cells were fixed with PAF 4% and treated as described in the materials and methods section. Selective permeabilization was done by using Triton X-100 or digitonin. Cells were labeled with antibodies against TMEM165 and GM130 before confocal microscopy visualization.

SPCA1 knockdown does not prevent TMEM165-induced degradation

Since the ELGDK motif is oriented toward the cytosol, we wished to assess whether TMEM165 responded to changes in cytosolic or Golgi luminal Mn2+. To tackle this point, we tested the contribution of SPCA1 (known to be one of the major Golgi Mn2+ importer) in the Mn2+-induced degradation of TMEM165. The impact of knockdown of SPCA1 on the Mn2+-induced degradation of TMEM165 was assessed. SiRNA of SPCA1 was very efficient as 85% of the protein was depleted compared with untreated cells. Interestingly, knockdown of SPCA1 did not abolish the Mn2+-induced degradation of TMEM165 (Figure 7). The results showed that TMEM165 loss exceeded 80% after 8 h of Mn2+ treatment both in siSPCA1 cells and untreated cells. This highly strengthens the fact that the degradation of TMEM165 is not dependent on Golgi luminal Mn2+ changes.

SPCA1knockdown does not prevent the Mn-induced degradation of TMEM165.

Figure 7.
SPCA1knockdown does not prevent the Mn-induced degradation of TMEM165.

(A) Steady-state cellular level of TMEM165 and SPCA1 in control and siSPCA1 HeLa cells. Control and siSPCA1 HeLa cells were cultured in the absence or presence of MnCl2 (500 µMn, 8 h); total cell lysates were then prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 and SPCA1 protein levels after normalization with actin (N = 3; ***P-value  < 0.001).

Figure 7.
SPCA1knockdown does not prevent the Mn-induced degradation of TMEM165.

(A) Steady-state cellular level of TMEM165 and SPCA1 in control and siSPCA1 HeLa cells. Control and siSPCA1 HeLa cells were cultured in the absence or presence of MnCl2 (500 µMn, 8 h); total cell lysates were then prepared, and subjected to SDS–PAGE and western blot with the indicated antibodies. Right panel shows the quantification of TMEM165 and SPCA1 protein levels after normalization with actin (N = 3; ***P-value  < 0.001).

Discussion

Our previous work has shown that the observed Golgi glycosylation defect due to a lack of Gdt1p/TMEM165 resulted from a Golgi Mn2+ homeostasis defect, then leading to strong Golgi glycosylation abnormalities. Interestingly, we demonstrated that such defects could totally be suppressed by manganese supplementation, strongly suggesting that TMEM165 could somehow be involved in the Golgi transport of Mn2+. It has been shown in yeast that Smf1p and Smf2p, members of the Nramp family of metal transporters, are tightly regulated by different intracellular Mn2+ concentrations [1820]. When cells are exposed to toxic Mn2+ concentrations, Smf1p and Smf2p are targeted to the vacuole for degradation, thus stopping the Mn2+ cellular entry. To test whether TMEM165 falls under the same regulation, TMEM165 stability for Mn2+ was tested. Our results showed that TMEM165 was highly sensitive to Mn2+ as manganese supplementation targets TMEM165 in the lysosomal degradation pathway. Although, intriguingly, the molecular mechanisms by which TMEM165 is degraded following Mn2+ exposure are currently not known, the Mn2+-induced degradation of Gdt1p-Myc in yeasts demonstrated that this mechanism is conserved during evolution. Another mammalian Golgi protein GPP130 has been reported to be sensitive to Mn2+ [16]. While the obtained results are very similar to the one observed for GPP130, several lines of evidences tend to prove that the molecular mechanisms could be different. First, the manganese sensitivity is different as 25 µM manganese is sufficient to engage TMEM165 in the lysosomal degradation pathway, while at this concentration GPP130 is stable. We cannot, however, avoid the fact that this observed difference in manganese sensitivity is coming from the different binding affinities of manganese for these two proteins. Second, the manganese-induced degradation rate of TMEM165 is faster than that of GPP130, as TMEM165 accumulation was never seen in punctate structures under Mn2+ supplementation. Because GPP130 and TMEM165 present high sensitivity to manganese, we cannot exclude a functional link between these two Golgi proteins.

Interestingly, while the R126H mutation remains Mn2+ responsive, the glutamic acid (E108) in the highly conserved ELGDK motif was shown to be insensitive to Mn2+ exposure and then crucial in TMEM165 Mn2+-induced degradation. One can suppose that these two mutations act differently on the Mn2+-induced degradation mechanisms of TMEM165. Our data also show that the E108G TMEM165 mutant form is able to rescue the glycosylation defect, although less efficiently than the wt-TMEM165 form. This suggests that the Mn2+-induced degradation mechanism is independent of the function of TMEM165 in Golgi glycosylation. According to the prediction of TMEM165 membrane topology, this motif is oriented towards the cytosol and located between the second and the third transmembrane domains of TMEM165. Our results show that TMEM165 responds to changes in cytosolic Mn2+ and not Golgi luminal changes. Although the R126H mutation reduces basal TMEM165 expression, the protein remains Mn responsive.

The other important question is, why is TMEM165 degraded by high cytosolic Mn2+ concentration? While we currently do not have the answer, our data raise several hypotheses. As a slight fraction of TMEM165 can be found at the plasma membrane, the degradation could be a mechanism to prevent Mn2+ entry through the plasma membrane. As mammalian cells can, however, transport the metal by other plasma membrane transporters, this hypothesis is not likely. When cells are exposed to high manganese concentrations, the plasma membrane transporters import the dangerous metal in the cytosol where it accumulates and impairs many fundamental cellular processes. The detoxification is then crucial to avoid the impairment of these processes. It is known that SPCA1, the Golgi P-type ATPase essential to import cytosolic Ca2+ but also Mn2+ inside the Golgi lumen, is the major way for eliminating the surplus of cytosolic Mn2+ from the cell. As TMEM165 is degraded when the manganese level becomes toxic, we can hypothesize that this mechanism participates in detoxification. This is still unclear how TMEM165 participates in such a process, but one can think that TMEM165, in the presence of high Mn2+ concentration in the Golgi, could transport back the Mn2+ into the cytosol. In that case, the specific degradation of TMEM165 in response to Mn2+ would prevent the Mn2+ from the Golgi to be recaptured back into the cytosol, a mechanism that would definitely annihilate the efforts made by SPCA1.

Overall, our studies highlight TMEM165 as a novel Golgi Mn2+ sensitive protein in mammalian cells. This discovery sheds light on a novel actor involved in the regulation of intracellular Mn2+ homeostasis and the pathophysiological mechanisms in TMEM165-CDG patients.

Material and methods

Antibodies and other reagents

Anti-TMEM165 and anti-β Actin antibodies were from Sigma–Aldrich (St Louis, MO, U.S.A.). The other anti-TMEM165 antibody was purchased from Thermo Fisher Scientific (Waltham, MA, U.S.A.). Anti-SPCA1 antibody was purchased from Abcam (Cambridge, U.K.). Anti-GM130 antibody was from BD Biosciences (Franklin lakes, NJ, U.S.A.). Anti-GPP130 antibody was purchased from Covance (Princeton, NJ, U.S.A.). Anti-myc (9E10) was purchased from Santa Cruz Biotechnology (Dallas, TX, U.S.A.). Goat anti-rabbit or goat anti-mouse immunoglobulins HRP-conjugated were purchased from Dako (Glostrup, Denmark). Polyclonal goat anti-rabbit or goat anti-mouse conjugated with Alexa Fluor were purchased from Thermo Fisher Scientific (Waltham, MA, U.S.A.). Manganese (II) chloride tetrahydrate was from Riedel-de-Haën (Seelze, Germany). All other chemicals were from Sigma–Aldrich unless otherwise specified.

Constructs, vector engineering and mutagenesis

Plasmids, pcDNA3.1 derivatives expressing either wt-TMEM165 or p.E108G version of TMEM165 (c.A323G mutation), have been generated by Ezyvec (Lille, France).

Generation of TMEM165 knockouts

HEK293T cells (ATTC) were grown in Dulbecco's Modified Eagle's Medium (DMEM)/F12 medium (Thermo Scientific) supplemented with 10% fetal bovine serum (FBS; Atlas Biologicals). Cells were maintained at 37° and 5% CO2 in a 90% humidified incubator. HEK293T TMEM165 stable knockouts were generated using the CRISPR technique [2124]. gRNA sequences were purchased from Genecopoeia (Catalog No. HCP214780-SG01-3). HEK293T cells were transfected with plasmid containing gRNA and a separate plasmid containing Cas9 and mCherry. HEK293 cells were transfected in a 6-well plate at 70% confluence using Lipofectamine 2000 in Opti-MEM (Thermo Scientific). Cells were incubated with the lipid-DNA complexes for 5 h after which the cells were supplemented with DMEM/F12 with FBS at a final concentration of 5%. The medium was changed to DMEM/F12 with 10% FBS 24 h after transfection. Twelve days after transfection, cells were single-cell sorted and knockout colonies were identified by immunofluorescence and western blot using antibodies to TMEM165 (Sigma). Sequencing of knockouts identified deletions in exon 1.

Cell culture and transfections

All cell lines were maintained in DMEM supplemented with 10% FBS (Lonza, Basel, Switzerland), at 37°C in a humidity-saturated 5% CO2 atmosphere. Transfections were performed using Lipofectamine 2000® (Thermo Scientific) according to the manufacturer's guidelines. For drug treatments, incubations were done as described in each figure.

Immunofluorescence staining

Cells were seeded on coverslips for 12–24 h, washed once in Dulbecco's Phosphate Buffer Saline (DPBS, Lonza) and fixed either with 4% paraformaldehyde (PAF) in PBS (pH 7.3) for 30 min at room temperature or with ice-cold methanol for 10 min at room temperature. Coverslips were then washed three times with PBS. Only if the fixation had been done with PAF, cells were permeabilized with 0.5% Triton X-100 in PBS for 15 min and then washed three times with PBS. Coverslips were then put in saturation for 1 h in blocking buffer [0.2% gelatin, 2% Bovine Serum Albumin (BSA), 2% FBS (Lonza) in PBS], followed by incubation for 1 h with primary antibody diluted at 1:100 in blocking buffer. After washing with PBS, cells were incubated for 1 h with Alexa 488-, Alexa 568-, or Alexa 700-conjugated secondary antibody (Life Technologies) diluted at 1:600 in blocking buffer. After washing three times with PBS, coverslips were mounted on glass slides with Mowiol. Fluorescence was detected through an inverted Zeiss LSM780 confocal microscope. Acquisitions were done using the ZEN pro 2.1 software (Zeiss, Oberkochen, Germany). For selective membrane permeabilization, we have used digitonin at 5 µg/ml. Stock solution was prepared at 5 mg/ml in absolute ethanol, 0.3 M sucrose, 0.1 M KCl, 2.5 mM MgCl2, 1 mM EDTA, 10 mM HEPES, pH 6.9. Permeabilization was done at 4°C for 15 min.

Image analyses

Immunofluorescence images were analyzed using TisGolgi, a home-made imageJ (http://imagej.nih.gov/ij) plugin developed by TISBio and available upon request. Basically, the program automatically detects and discriminates Golgi and vesicles, based on morphological parameters such as size and sphericity. Then, the program calculates for each image the number of detected objects, their size and mean fluorescence intensity. Co-localization analyses were done using JACoP plugin and performed according to the guidelines suggested by Bolte et al. [25].

Western blotting

Cells were scraped in DPBS and then centrifuged at 5000×g for 3 min. Supernatant was discarded and cells were then resuspended in RIPA buffer [Tris/HCl 50 mM (pH 7.9), NaCl 120 mM, NP40 0.5%, EDTA 1 mM, Na3VO4 1 mM, NaF 5 mM] supplemented with a protease cocktail inhibitor (Roche Diagnostics, Penzberg, Germany). Cell lysis was done by passing the cells several times through a syringe with a 26G needle. Cells were centrifuged for 30 min at 20 000×g. The supernatant containing protein was estimated with the Micro BCA Protein Assay Kit (Thermo Scientific). A 20 µg aliquot of total protein lysate was put in NuPAGE LDS sample buffer (Invitrogen) (pH 8.4) supplemented with 4% β-mercaptoethanol (Fluka). Samples were heated for 10 min at 95°C, then separated on 4–12% Bis–Tris gels (Invitrogen) and transferred to nitrocellulose membrane Hybond ECL (GE Healthcare, Little Chalfont, U.K.). The membranes were blocked in blocking buffer (5% milk powder in TBS-T [1X TBS with 0.05% Tween20]) for 1 h at room temperature, then incubated overnight with the primary antibodies (used at a dilution of 1:1000, except for anti-myc, used at 1:200) in blocking buffer, and washed three times for 5 min in TBS-T. The membranes were then incubated with the peroxidase-conjugated secondary goat anti-rabbit or goat anti-mouse antibodies (Dako; used at a dilution of 1:10 000) in blocking buffer for 1 h at room temperature and later washed three times for 5 min in TBS-T. Signal was detected with chemiluminescence reagent (ECL 2 Western Blotting Susbtrate, Thermo Scientific) on imaging film (GE Healthcare, Little Chalfont, U.K.).

Yeast strains, media and lysis

Yeast strains originating from BY4741 background were used for the experiments (gdt1Δ: Mata his3Δ1 leu2Δ0 ura3Δ0 gdt1Δ::KanMX4). Yeasts were cultured at 30°C. Cultures in liquid media are done under light shaking. Rich medium, named YEP medium, contains yeast extract (10 g/l, Difco), Bacto-peptone (20 g/l, Difco); YPD medium is a YEP medium supplemented with 2% d-glucose (Sigma–Aldrich). Before any analysis, a preculture in YPD medium is done and a volume equivalent to 10 OD600 nm unit is transferred into a bigger volume of YPD medium. Culture begins at a volume equivalent to 6 OD600 nm unit until 18 OD600 nm. MnCl2 was added at this step at the indicated concentration and yeasts were harvested at the indicated times. Yeasts were centrifuged for 5 min at 3500×g. The supernatant was discarded and the pellet was kept frozen at −20°C. Yeast lysis was performed as described by Ballou et al. [26]. Western blot experiments were done as described above.

Cell surface biotinylation

Cells were plated to reach 70–80% confluence on the day of the experiment. Cells were kept on ice all the time. Cells were washed four times with PBS+/+ (containing Ca and Mg), pH8. An aliquot of 1.5 ml PBS+/+ (pH 8) with 7.5 µl biotin was added per dish [Biotin: EZ Link Sulfo-NHS-SS-Biotin (Life Technologies, Carlsbad, CA, U.S.A.), final concentration of 0.5 mg/ml in DMSO]. Cells were then incubated 30 min in a cold room on slow rocking and then washed three times with PBS+/+ (pH8). Cells were quenched 15 min with 1.5 ml of PBS+/+ glycine 100 mM, BSA 0.5% in a cold room on slow rocking and then washed three times with PBS+/+ glycine. Cells were scraped in 200 µl of lysis buffer [50 mM HEPES (pH 7.2), 100 mM NaCl, 1% Triton X-100, protease inhibitors], incubated 20 min on ice and centrifuged for 15 min at 20 000×g at 4°C. The supernatant was kept and the protein concentration was measured. For the pull-down, put the maximum amount of protein, ideally 500 µg in 1 ml final (lysis buffer) + 30 µl streptavidin beads. Incubate 4 h at 4°C on a wheel and then centrifuge at 4000×g for 1 min at 4°C. Wash three times with 1 ml of lysis buffer (not supplemented with protease inhibitors) and mix well by inverting the tubes 30 times. Centrifuge at 4000×g for 1 min at 4°C and remove the supernatant with a flat end tip. Add NuPAGE LDS sample buffer (Invitrogen) pH 8.4 supplemented with 4% β-mercaptoethanol (Fluka). Samples are boiled at 70°C for 10 min (do not boil if one wishes to reveal TMEM165 on western blot afterwards) and then centrifuged at 1000×g for 1 min, and the supernatant was collected. Samples were frozen at −20°C. Samples are ready to load on gel.

Statistical analysis

Comparisons between groups were performed using Student's t-test for two variables with equal or different variances, depending on the result of the F-test.

Abbreviations

     
  • BSA

    bovine serum albumin

  •  
  • CDG

    congenital disorders of glycosylation

  •  
  • DMEM

    Dulbecco's Modified Eagle's Medium

  •  
  • DPBS

    Dulbecco's Phosphate Buffer Saline

  •  
  • FBS

    fetal bovine serum

  •  
  • PAF

    paraformaldehyde

  •  
  • SPCA1

    secretory pathway Ca-ATPase 1

  •  
  • wt

    wild-type

Author Contribution

F.F. obtained financial support to design this study and he wrote the paper. F.F. and S.P. coordinated the study. S.P. and E.D. performed and analyzed most of the experiments. L.C. and V.L. performed the CRISPR-cas9 TMEM165 cells. M.H. performed the experiment in Figure 7. C.S. provided technical assistance for colocalization studies. D.V. performed immunofluorescence microscopy experiments. E.L. and A.K. performed and analyzed the experiment in Figure 4C. R.P. provided technical assistance on surface biotinylation. S.D., W.M. and M.A.K. provided advices. G.D.B. and P.M. reviewed the yeast results. G.M. reviewed the paper and provided us the R126H deficient CDG patients. T.M. provided us the E108G TMEM165 deficient patients.

Funding

This work was supported by the French National Research Agency [SOLV-CDG to F.F] and the Mizutani Foundation for Glycoscience [to F.F.] and EURO-CDG-2 that has received funding from the European Union’s Horizon 2020 research and innovation program under the ERA-NET Cofund action N° 643578. V.L. was supported by the NIH grants [GM083144 and U54 GM105814].

Acknowledgments

We are indebted to Dr Dominique Legrand of the Research Federation FRABio (Univ. Lille, CNRS, FR 3688, FRABio, Biochimie Structurale et Fonctionnelle des Assemblages Biomoléculaires) for providing the scientific and technical environment conducive to achieving this work. We thank the BioImaging Center of Lille, especially Christian Slomianny and Elodie Richard, for the use of the Leica LSM70.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Au
,
C.
,
Benedetto
,
A.
and
Aschner
,
M.
(
2008
)
Manganese transport in eukaryotes: the role of DMT1
.
Neurotoxicology
29
,
569
576
doi:
2
Perl
,
D.P.
and
Olanow
,
C.W.
(
2007
)
The neuropathology of manganese-induced Parkinsonism
.
J. Neuropathol. Exp. Neurol.
66
,
675
682
doi:
3
Squitti
,
R.
,
Gorgone
,
G.
,
Panetta
,
V.
,
Lucchini
,
R.
,
Bucossi
,
S.
,
Albini
,
E.
et al. 
(
2009
)
Implications of metal exposure and liver function in Parkinsonian patients resident in the vicinities of ferroalloy plants
.
J. Neural. Transm.
116
,
1281
1287
doi:
4
Park
,
J.H.
,
Hogrebe
,
M.
,
Grüneberg
,
M.
,
DuChesne
,
I.
,
Heiden
,
A.L.
von der Reunert
,
J.
et al. 
(
2015
)
SLC39A8 deficiency: a disorder of manganese transport and glycosylation
.
Am. J. Hum. Genet.
97
,
894
903
doi:
5
Potelle
,
S.
,
Morelle
,
W.
,
Dulary
,
E.
,
Duvet
,
S.
,
Vicogne
,
D.
,
Spriet
,
C.
et al. 
(
2016
)
Glycosylation abnormalities in Gdt1p/TMEM165 deficient cells result from a defect in Golgi manganese homeostasis
.
Hum. Mol. Genet.
25
,
1489
1500
doi:
6
Garrick
,
M.D.
,
Dolan
,
K.G.
,
Horbinski
,
C.
,
Ghio
,
A.J.
,
Higgins
,
D.
,
Porubcin
,
M.
et al. 
(
2003
)
DMT1: a mammalian transporter for multiple metals
.
Biometals
16
,
41
54
doi:
7
Forbes
,
J.R.
and
Gros
,
P.
(
2003
)
Iron, manganese, and cobalt transport by Nramp1 (Slc11a1) and Nramp2 (Slc11a2) expressed at the plasma membrane
.
Blood
102
,
1884
1892
doi:
8
Chen
,
P.
,
Bowman
,
A.B.
,
Mukhopadhyay
,
S.
and
Aschner
,
M.
(
2015
)
SLC30A10: a novel manganese transporter
.
Worm
4
,
e1042648
doi:
9
Chen
,
P.
,
Chakraborty
,
S.
,
Mukhopadhyay
,
S.
,
Lee
,
E.
,
Paoliello
,
M.M.B.
,
Bowman
,
A.B.
et al. 
(
2015
)
Manganese homeostasis in the nervous system
.
J. Neurochem.
134
,
601
610
doi:
10
Girijashanker
,
K.
,
He
,
L.
,
Soleimani
,
M.
,
Reed
,
J.M.
,
Li
,
H.
,
Liu
,
Z.
et al. 
(
2008
)
Slc39a14 gene encodes ZIP14, a metal/bicarbonate symporter: similarities to the ZIP8 transporter
.
Mol. Pharmacol.
73
,
1413
1423
doi:
11
Dode
,
L.
,
Andersen
,
J.P.
,
Vanoevelen
,
J.
,
Raeymaekers
,
L.
,
Missiaen
,
L.
,
Vilsen
,
B.
et al. 
(
2006
)
Dissection of the functional differences between human secretory pathway Ca2+/Mn2+-ATPase (SPCA) 1 and 2 isoenzymes by steady-state and transient kinetic analyses
.
J. Biol. Chem.
281
,
3182
3189
doi:
12
He
,
W.
and
Hu
,
Z.
(
2012
)
The role of the Golgi-resident SPCA Ca2+/Mn2+ pump in ionic homeostasis and neural function
.
Neurochem. Res.
37
,
455
468
doi:
13
Mukhopadhyay
,
S.
and
Linstedt
,
A.D.
(
2011
)
Identification of a gain-of-function mutation in a Golgi p-type ATPase that enhances Mn2+ efflux and protects against toxicity
.
Proc. Natl Acad. Sci. U.S.A.
108
,
858
863
doi:
14
Foulquier
,
F.
,
Amyere
,
M.
,
Jaeken
,
J.
,
Zeevaert
,
R.
,
Schollen
,
E.
,
Race
,
V.
et al. 
(
2012
)
TMEM165 deficiency causes a congenital disorder of glycosylation
.
Am. J. Hum. Genet.
91
,
15
26
doi:
15
Masuda
,
M.
,
Braun-sommargren
,
M.
,
Crooks
,
D.
and
Smith
,
D.R.
(
2013
)
Golgi phosphoprotein 4 (GPP130) is a sensitive and selective cellular target of manganese exposure
.
Synapse
67
,
205
215
doi:
16
Mukhopadhyay
,
S.
,
Bachert
,
C.
,
Smith
,
D.R.
and
Linstedt
,
A.D.
(
2010
)
Manganese-induced trafficking and turnover of the cis-Golgi glycoprotein GPP130
.
Mol. Biol. Cell
21
,
1282
1292
doi:
17
Schulte Althoff
,
S.
,
Grüneberg
,
M.
,
Reunert
,
J.
,
Park
,
J.H.
,
Rust
,
S.
,
Mühlhausen
,
C.
et al. 
(
2015
)
TMEM165 deficiency: postnatal changes in glycosylation
.
JIMD Rep.
26
,
21
29
doi:
18
Culotta
,
V.C.
,
Yang
,
M.
and
Hall
,
M.D.
(
2005
)
Manganese transport and trafficking: lessons learned from Saccharomyces cerevisiae
.
Eukaryot. Cell
4
,
1159
1165
doi:
19
García-Rodríguez
,
N.
,
Manzano-López
,
J.
,
Muñoz-Bravo
,
M.
,
Fernández-García
,
E.
,
Muñiz
,
M.
and
Wellinger
,
R.E.
(
2015
)
Manganese redistribution by calcium-stimulated vesicle trafficking bypasses the need for P-type ATPase function
.
J. Biol. Chem.
290
,
9335
9347
doi:
20
Reddi
,
A.R.
,
Jensen
,
L.T.
and
Culotta
,
V.C.
(
2009
)
Manganese homeostasis in Saccharomyces cerevisiae
.
Chem. Rev.
109
,
4722
4732
doi:
21
Bailey Blackburn
,
J.
,
Pokrovskaya
,
I.
,
Fisher
,
P.
,
Ungar
,
D.
and
Lupashin
,
V.V.
(
2016
)
COG complex complexities: detailed characterization of a complete set of HEK293T cells lacking individual COG subunits
.
Front. Cell Dev. Biol.
4
,
23
doi:
22
Cong
,
L.
,
Ran
,
F.A.
,
Cox
,
D.
,
Lin
,
S.
,
Barretto
,
R.
,
Habib
,
N.
et al. 
(
2013
)
Multiplex genome engineering using CRISPR/Cas systems
.
Science
339
,
819
823
doi:
23
Jinek
,
M.
,
Chylinski
,
K.
,
Fonfara
,
I.
,
Hauer
,
M.
,
Doudna
,
J.A.
and
Charpentier
,
E.
(
2012
)
A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity
.
Science
337
,
816
821
doi:
24
Mali
,
P.
,
Yang
,
L.
,
Esvelt
,
K.M.
,
Aach
,
J.
,
Guell
,
M.
,
DiCarlo
,
J.E.
et al. 
(
2013
)
RNA-guided human genome engineering via Cas9
.
Science
339
,
823
826
doi:
25
Bolte
,
S.
and
Cordelières
,
F.P.
(
2006
)
A guided tour into subcellular colocalization analysis in light microscopy
.
J. Microsc.
224
,
213
232
doi:
26
Ballou
,
C.E.
(
1990
)
[36] isolation, characterization, and properties of Saccharomyces cerevisiae mnn mutants with nonconditional protein glycosylation defects
.
Methods. Enzymol.
185
,
440
470
doi:

Author notes

These authors contributed equally to this work.